Dongdong Yu

Dongdong Yu

Auto-generated profile, based on publicly available information

Dongdong Yu

The Second Clinical College of Guangzhou University of Chinese Medicine, Guangzhou, Guangdong Province, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 November 2025 in Journal of Medical Internet Research
Verified
Journal Article
Open access
XZXiaoying Zhong
Xiaoying Zhong
SLSiyi Li
Siyi Li
ZCZhao Chen
Zhao Chen
LGLong Ge
Long Ge
DYDongdong Yu
Dongdong Yu
HSHongcai Shang
Hongcai Shang
Background: The application of large language models (LLMs) in health care holds significant potential for enhancing patient care and advancing medical research. However, the protection of patient privacy remains a critical issue, especially when handling patient health information (PHI). Objective: This scoping review aims to evaluate the adequacy of current approaches and identify areas in need of improvement to ensure robust patient privacy protection in the existing studies about PHI-LLMs within the health care domain. Methods: A search of the literature published from January 1, 2022, to July 20, 2025, was performed on July 20, 2025, using 2 databases (PubMed and Embase). This scoping review focused on the following three research questions: (1) What studies on the development and application of LLMs using PHI currently exist within the health care domain? (2) What patient privacy considerations are addressed in existing PHI-LLMs research, and are these measures sufficient? (3) How can future research on the development and application of LLMs using PHI better protect patient privacy? Studies were included if they focused on the development and application of LLMs within health care using PHI, encompassing activities such as model construction, fine-tuning, optimization, testing, and performance comparison. Eligible literature comprised original research articles written in English. Conversely, studies were excluded if they used publicly available datasets, under the assumption that such data have been adequately deidentified. Additionally, non-English publications, reviews, abstracts, incomplete reports, and preprints were excluded from the review due to the lack of rigorous peer review. Results: This study systematically identified 9823 studies on PHI-LLM and included 464 studies published between 2022 and 2025. Among the 464 studies, (1) a small number of studies neglected ethical review (n=45, 9.7%) and patient informed consent (n=148, 31.9%) during the research process, (2) more than a third of the studies (n=178, 38.4%) failed to report whether to implement effective measures to protect PHI, and (3) there was a significant lack of transparency and comprehensive detail in anonymization and deidentification methods. Conclusions: We propose comprehensive recommendations across 3 phases—study design, implementation, and reporting—to strengthen patient privacy protection and transparency in PHI-LLM. This study emphasizes the urgent need for the development of stricter regulatory frameworks and the adoption of advanced privacy protection technologies to effectively safeguard PHI. It is anticipated that future applications of LLMs in the health care field will achieve a balance between innovation and robust patient privacy protection, thereby enhancing ethical standards and scientific credibility.

20 August 2024 in Expert Opinion on Drug Safety
Verified
Journal Article
Closed access
DYDongdong Yu
Dongdong Yu
TCTing Cheng
Ting Cheng
TLTong Liu
Tong Liu
WXWenjun Xu
Wenjun Xu
DLDawei Liu
Dawei Liu
XCXiaoyu Cheng
Xiaoyu Cheng
The use of proteasome inhibitors (PIs), namely Bortezomib and Carfilzomib, revolutionized multiple myeloma (MM) treatment. Understanding their distinct adverse event (AE) profiles aids in tailored treatment plans. We analyzed FDA Adverse Event Reporting System (FAERS) data (Q1 2012-Q4 2023) for Bortezomib and Carfilzomib, utilizing reporting odds ratio (ROR), proportional reporting ratio (PRR), and Bayesian confidence propagation neural network (BCPNN). FAERS yielded 19,720 Bortezomib and 12,252 Carfilzomib AE reports. Males aged 45-65 exhibited higher AE susceptibility. Common AE systems included Infections, Nervous System Disorders, Blood Disorders, General Disorders, Cardiac Disorders, and Renal Disorders. New Bortezomib signals were sepsis and colitis. Carfilzomib exhibited elevated cardiac and renal toxicity but reduced peripheral neuropathy and thrombocytopenia. FAERS analysis revealed new AE signals (sepsis, colitis) for Bortezomib and highlighted Carfilzomib's heightened cardiac and renal risks compared to Bortezomib. Balancing PIs' benefits and risks is crucial for clinical decision-making.

7 September 2025 in Food Science & Nutrition
Verified
Journal Article
Open access
DYDongdong Yu
Dongdong Yu
TCTing Cheng
Ting Cheng
YSYuxuan Shi
Yuxuan Shi
XQXingying Qiu
Xingying Qiu
GLGeng Li
Geng Li
ZWZehuai Wen
Zehuai Wen
The relationship between gut microbiota, diet, and cardiovascular‐kidney‐metabolic (CKM) health has attracted attention. However, the relationship between the dietary index for gut microbiota (DI‐GM) and CKM syndrome has not yet been studied. Patients diagnosed with CKM syndrome from the NHANES 2007–2018 data were included. Dietary recall data were used to calculate DI‐GM. Restricted cubic splines (RCS) were employed to explore nonlinear relationships and determine the threshold for DI‐GM. The relationship between DI‐GM and CKM syndrome was analyzed using weighted logistic regression. Further, potential mediating roles of phenotype age acceleration (PAA), biological age acceleration (BAA), body mass index (BMI), and body roundness index (BRI) were explored. Sensitivity analysis using inverse probability of treatment weighting (IPTW) was also conducted. A total of 7252 participants were included, with the high DI‐GM group as the reference. In the crude model, the risk of CKM syndrome in the low DI‐GM group was significantly higher (OR = 1.27, 95% CI = 1.09, 1.49, p < 0.05). This association remained significant in the fully adjusted model (OR = 1.34, 95% CI = 1.10, 1.64, p < 0.05). The results of the IPTW analysis were consistent with the above findings. In the association between DI‐GM and CKM syndrome, significant mediating effects were observed for PAA (mediated proportion (MP): 14.84%, p < 0.001), BAA (MP: 21.45%, p < 0.001), BMI (MP: 24.17%, p < 0.001), and BRI (MP: 35.85%, p < 0.001). Low DI‐GM is associated with an increased prevalence of CKM syndrome. PAA, BAA, BRI, and BMI significantly mediate the relationship between DI‐GM and CKM syndrome.

3 July 2026 in Medicine
Verified
Journal Article
Open access
DYDongdong Yu
Dongdong Yu
YSYuxuan Shi
Yuxuan Shi
HFHuaying Fu
Huaying Fu
TLTong Liu
Tong Liu
YWYongkang Wang
Yongkang Wang
SWSihai Wang
Sihai Wang
Evidence regarding the mortality risk associated with physical activity (PA) and multimorbidity is limited. This study investigates trends in the prevalence of multimorbidity in the population and the mortality risk associated with PA and multimorbidity. The association between PA and overall survival was analyzed in 39,940 subjects from the National Health and Nutrition Examination Survey cohorts. Among the 39,940 participants (weighted average age: 46.84; 50.86% female), the age-standardized prevalence of multimorbidity rose significantly, from 15.87% (95% confidence interval [CI] = 14.86%–16.87%) in 1999 to 2000 to 19.96% (95% CI = 17.61%–22.32%) in 2017 to 2018. The population’s multimorbidity detection rate was 25.22%. The all-cause mortality risks for multimorbidity were: dual-multimorbidity (hazard ratio [HR] = 1.97, 95% CI = 1.76–2.20), triple-multimorbidity (HR = 2.69, 95% CI = 2.35–3.08), quadruple-multimorbidity (HR = 3.22, 95% CI = 2.59–3.99). “Diabetes mellitus + cardiovascular disease + respiratory diseases” had the highest mortality risk (HR = 4.37, 95% CI = 2.67–7.16). PA reduced all-cause and cause-specific mortality, with moderate-intensity PA lowering all-cause mortality by 22% compared to low-intensity PA, and high-intensity PA by 6% compared to moderate-intensity PA. Multimorbidity led to a 5.64 years life expectancy loss, while PA could recover up to 2.31 years. PA potentially recouped 2.96 and 2.19 years of life lost due to comorbidity in females and males, respectively. According to National Health and Nutrition Examination Survey data from 1999 to 2000 to 2017 to 2018, multimorbidity prevalence increased. PA reduces all-cause mortality by 32%. Greater PA intensity correlates with more significant mortality reduction. Multimorbidity cuts life expectancy by 5.64 years, recoverable by PA (2.31 years). Females regain 2.96 years; males, 2.19 years.

26 April 2025 in JMIR Preprints
Verified
Preprint
Open access
XZXiaoying Zhong
Xiaoying Zhong
SLSiyi Li
Siyi Li
ZCZhao Chen
Zhao Chen
LGLong Ge
Long Ge
DYDongdong Yu
Dongdong Yu
HSHongcai Shang
Hongcai Shang
BACKGROUND The application of large language models (LLMs) in health care holds significant potential for enhancing patient care and advancing medical research. However, the protection of patient privacy remains a critical issue, especially when handling patient health information (PHI). OBJECTIVE This scoping review aims to evaluate the adequacy of current approaches and identify areas in need of improvement to ensure robust patient privacy protection in the existing studies about PHI-LLMs within the health care domain. METHODS A search of the literature published from January 1, 2022, to July 20, 2025, was performed on July 20, 2025, using 2 databases (PubMed and Embase). This scoping review focused on the following three research questions: (1) What studies on the development and application of LLMs using PHI currently exist within the health care domain? (2) What patient privacy considerations are addressed in existing PHI-LLMs research, and are these measures sufficient? (3) How can future research on the development and application of LLMs using PHI better protect patient privacy? Studies were included if they focused on the development and application of LLMs within health care using PHI, encompassing activities such as model construction, fine-tuning, optimization, testing, and performance comparison. Eligible literature comprised original research articles written in English. Conversely, studies were excluded if they used publicly available datasets, under the assumption that such data have been adequately deidentified. Additionally, non-English publications, reviews, abstracts, incomplete reports, and preprints were excluded from the review due to the lack of rigorous peer review. RESULTS This study systematically identified 9823 studies on PHI-LLM and included 464 studies published between 2022 and 2025. Among the 464 studies, (1) a small number of studies neglected ethical review (n=45, 9.7%) and patient informed consent (n=148, 31.9%) during the research process, (2) more than a third of the studies (n=178, 38.4%) failed to report whether to implement effective measures to protect PHI, and (3) there was a significant lack of transparency and comprehensive detail in anonymization and deidentification methods. CONCLUSIONS We propose comprehensive recommendations across 3 phases—study design, implementation, and reporting—to strengthen patient privacy protection and transparency in PHI-LLM. This study emphasizes the urgent need for the development of stricter regulatory frameworks and the adoption of advanced privacy protection technologies to effectively safeguard PHI. It is anticipated that future applications of LLMs in the health care field will achieve a balance between innovation and robust patient privacy protection, thereby enhancing ethical standards and scientific credibility.

30 December 2025 in PLOS ONE
Verified
Journal Article
Open access
DYDongdong Yu
Dongdong Yu
TCTing Cheng
Ting Cheng
YYYue Yang
Yue Yang
XQXingying Qiu
Xingying Qiu
GLGeng Li
Geng Li
ZWZehuai Wen
Zehuai Wen
This study explored the non-linear relationship between pan-immune inflammation value (PIV) and probable depression using NHANES data (2005–2018, n = 27,049). Restricted cubic spline analysis identified an optimal PIV cutoff (5.74). Weighted logistic regression revealed that individuals with PIV > 5.74 had 14% higher risk of probable depression (OR=1.14, 95% CI:1.02–1.28). Gender-stratified analyses showed significant PIV and probable depression associations in females (OR=1.20, 95% CI:1.03–1.40) but not in males. In males, higher PIV interacted with age, education, and alcohol consumption to influence probable depression risk (interaction P < 0.05), whereas no such interactions occurred in females. The findings suggest PIV may serve as a sex-specific biomarker for probable depression risk, with more stable associations in females. Further research is needed to elucidate the mechanisms underlying these gender differences and validate these findings before any clinical application.